Despite the recent substantial increase in data about cocaine's basic neurochemical mechanisms of action, progress towards the development of an effective pharmacological treatment for cocaine abuse has been disappointing. Our cocaine self-administration laboratory procedure is relatively more rapid and less expensive than a clinical trial, while still including the most salient feature of drug abuse-drug taking behavior. The procedure combines more traditional measures (e.g, subjective effects, cocaine craving, physiological measures) with a measure of actual cocaine use. We are proposing to use our laboratory model of repeated dose cocaine choice/self-administration to assess the potential efficacy of the D1 dopamine receptor antagonist, SCH 39166, for the treatment of cocaine dependence. Non-treatment seeking cocaine abusers will live on the hospital clinical research unit and will be maintained on placebo and SCH 3916, each for 8 days. Subjects will participate in 8 laboratory sessions in which they will have the opportunity to choose between repeated doses of smoked cocaine and $5. In addition to measuring their choice behavior, we will measure their cardiovascular and subjective responsivity to cocaine craving, and memory. The following questions will be addressed: (1) Does this medication interfere with the reinforcing efficacy of cocaine? This will be addressed by comparing cocaine self-administration following maintenance on SCH 391666 compared to maintenance on a placebo, (2) Does SCH 39166 interact with or disrupt other stimulus functions of cocaine? This will be addressed by comparing the effects of SCH 39166 and placebo on cocaine's subjective and physiological effects.